Join our community of smart investors

Deciphering mixed messages on equities

There’s a danger that the fall in US equities so far this year could turn into something very nasty, because when over-valued markets start to fall they can drop a lot. When the tech bubble burst in 2000 for example, the S&P 500 lost more than 40 per cent.

We do, though, have two warning signs that can help us avoid such losses.

One is whether the index is below its 10-month (or 200-day) moving average. Mebane Faber at Cambria Investment Management has shown that a rule of selling when the S&P is below this average would have given better risk-adjusted returns over the long run than a simple buy-and-hold strategy.

The rule has worked well in recent years. Since 1990 the US market has risen by an average of 8.7 per cent in real terms in the 12 months after the index has been above its 10-month average, but by only 3.7 per cent after it has been below it. And the rule would have saved us from some huge losses. It would have got us out of the market for most of the period from September 2000 to March 2003 during which time the market fell 42 per cent, and from November 2007 to May 2009 during which time it fell 38 per cent.

Of course, the rule doesn’t always work so well. It told us to get out in March 2020 after the start of the pandemic had clobbered global markets, for example. We would therefore have missed out on the start of a big bounce back.

Which is a general problem with the rule. Although it does a great job of protecting us from long bear markets such as those of 2000-03 and 2007-09 it fails when a strategy of buying on dips works. It fails when the market is dominated by value investors (those who buy on dips), but works when it is dominated by momentum investors – those who sell because others have sold.

As I write, the rule is telling us to sell US stocks. This would serve us well if we get a long deflation of valuations like we saw in 2000, or if sky-high oil prices lead to recession, but not necessarily if the Russian war ends soon and well.

It, however, is not the only lead indicator of returns we have. There’s another: the shape of the yield curve, which I’ll define for these purposes as the gap between 10-year Treasury yields and the Fed funds rate. Since 1990 US stocks have risen by an average of 9.1 per cent in real terms in the 12 months after 10-year yields have been above the Fed funds rate, but they’ve fallen by an average of 3.9 per cent after they have been below it. Inverted yield curves in 2000 and 2007, for example, both led to big falls, while the bull market of 2009-19 followed mostly upward-sloping curves.

Right now, the curve is telling us to stay in equities, contradicting the message of the 10-month rule.

But this rule, like the 10-month rule, isn’t perfect. It told us to get out of equities in 2006 and early 2020, which would have meant missing out on months of nice gains, and to get into the market in mid-2008 before some heavy losses.

So, which rule do we believe? The answer is: both. A combination of the two has worked better than either in isolation.

We can quantify this. Since 1990 US equities have risen by an average of 10 per cent in real terms in the 12 months after both the 10-month rule and yield curve told us to buy. But it has fallen by an average of 6.8 per cent in the 12 months after both told us to sell. When the 10-month rule told us to buy but the yield curve to sell, the market has dropped by an average of 2.6 per cent. And when the yield curve has told us to buy but the 10-month rule to sell the market has risen by an average of 5.8 per cent.

With the latter being the message right now, investors therefore have reason for hope. Except, that is, for two things.

One is that there is variation about this average. Sometimes the 10-month sell signal has been correct while the yield curve buy signal has been wrong: this was the case in 2008, for example. Even with the best lead indicators we cannot time the market perfectly, any more than we can pick stocks perfectly.

The other is that we’ve reason to doubt the message of the yield curve now. It has worked well in the past because it has embodied the wisdom of crowds. Each investor has a little dispersed and fragmentary insight into where interest rates and the economy are heading. The yield curve aggregates together all these insights such that when 10-year yields are below short-term rates it really is a predictor of recession. The whole is much greater than the parts.

Right now, though, the main determinant of where the market is heading lies in the mind of Vladimir Putin. There’s no great reason to suppose that investors’ insights into this – even in aggregate – are as good as their opinions of the domestic economy. Which, I think, argues for discounting the message of the yield curve at least a little.

Personally, then, I am cautious about the market: I have around a 50 per cent cash weighting. Such caution might be mistaken. But I’d much rather be wrong in this direction than the other.